[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: spot vs line analyzer for telecine

jpb wrote:

> Would it be true to say that --in theory--,
> the 'dot' illumination of flying spot analysers --which only
> puts light onto one pixel of the film at a time-- can produce
> cleaner images than the 'line' illumination used for CCD
> imagers in which all the clear/transparent areas of the full
> scanned line can be detrimental (through the unavoidable parasitic
> reflexions in the optical path) to their opaque neighbours purity?
> ... or am I out to lunch?
> --jp
> ______________________________________________

I'm not sure why you say "analyzers", but I will presume you
are referring to scanners nevertheless.  Some people would
love to believe that the flying spot is small and precise enough
to illuminate only "one pixel of the film at a time".  Unfortunately,
such precision is not reality in the real world of film scanning
due to a number of characteristics surrounding the flying spot
(no pun intended) that keep it from being that perfect.

The spot is not perfectly round to begin with, nor is the spot
"on one pixel at one time" due to its limit of physical size and
phosphor persistence characteristics which is a limit to resolution,
and unfortunately also varies with age.  Both phosphor spot size
and persistence are natural limits of capable resolution, and
are directly related to scanning speed.  As phosphor scanning
speed increases, relative resolution decreases.

Cleaner images?  Although cleaner could mean a number of
subjective descriptions, probably most often it refers to lack of
noise, in the practical sense, in which case you need to consider
the following:

Flying spot phosphors produce their light predominantly in the
green part of the spectrum.  The red, and especially blue
components of their light output are substantially lower than
that of the green. This places a heavy dependency on an extremely
high gain, noise free detection system which is not so easy
to achieve, again in practicality.

The deposited phosphor particles that form the CRT screen
face has a granularity characteristic whereas the CCD imager
does not.  Even though this effect is very minute, it is enough
to cause a graininess effect particularly when film densities
are such to necessitate high amplification of the images optical
path.  The CCD pixel light uniformity is not super precise enough
(yet) and does have some pixel to pixel light variation, but it
can be compensated for much easier than phosphor grain.

What typically makes the CCD imager cleaner is its relatively
noise free, high resolution image.  Even though the amount of
light directed through the film to each of the CCD pixel elements
is far greater than that from a flying spot source, keep in mind that
the image is focused to each of those pixel sensors which are
typically much smaller than the flying spot size itself, and fixed
in size throughout their life.

Light spill or "parasitic reflections", as you say, are not
"unavoidable" with proper optical designs.  In fact, there are
more advantages than disadvantages with CCD imaging relating
to the optical path residual artifacts but, scan rate and anti-
aliasing are factors with CCD imaging.  The very fact that the
flying spot is not perfectly sized to down to each (relative)
pixel is a natural filter in itself.

There are also other variables to the design of a CCD scanner,
than just the fact that a CCD sensor is being used, that can cause
the image to appear like a very harsh video camera.  Although,
thanks to advancements in technology, CCD scanners today are
not anywhere near the same as they were even 4 years ago.

Michael C. Kaye

Now go to lunch... and bring me back a Diet Coke or something.....

	Thanks to Kat Dalton for support of the TIG
      TIG subscriber count is 846 on Mon Sep  8 12:12:04 PDT 1997
   archives and much more at http://www.alegria.com/telecinehome.html
     mailinglist digest available.... unsubscribe via a message to
        'telecine-request at alegria.com' with Subject: unsubscribe